The emergence of acid ceramidase as a therapeutic target for acute myeloid leukemia

Abstract

Introduction

Acute myeloid leukemia (AML) is the most common adult leukemia. Only a fraction of AML patients will survive with existing chemotherapy regimens. Hence, there is an urgent and unmet need to identify novel targets and develop better therapeutics in AML. In the past decade, the field of sphingolipid metabolism has emerged into the forefront of cancer biology due to its importance in cancer cell proliferation and survival. In particular, acid ceramidase (AC) has emerged as a promising therapeutic target due to its role in neutralizing the pro-death effects of ceramide.

Areas covered

This review highlights key information about AML biology as well as current knowledge on dysregulated sphingolipid metabolism in cancer and AML. We describe AC function and dysregulation in cancer, followed by a review of studies that report elevated AC in AML and compounds known to inhibit the enzyme.

Expert opinion

AML has a great need for new drug targets and better therapeutic agents. The finding of elevated AC in AML supports the concept that this enzyme represents a novel and realistic therapeutic target for this common leukemia. More effort is needed towards developing better AC inhibitors for clinical use and combination treatment with existing AML therapies.

1. Background and introduction

1.1. Acute myeloid leukemia

Acute myeloid leukemia (AML) is a malignancy of the blood and bone marrow characterized by uncontrolled proliferation of immature cells of the myeloid lineage. These cells, called blasts, overpopulate the blood as well as bone marrow and compromise production of normal blood cells. As a result, patients may present with symptoms of anemia, bleeding or infection. As an acute disease, AML exhibits sudden onset and rapid progression. Only 27% of patients survive five years past diagnosis¹. Prognosis is even worse for some groups, including those above the age of 65 and those whose disease has developed from myelodysplastic syndrome or following prior chemotherapy^2-4.

In addition to the challenge of a quickly progressing disease, AML is also extremely heterogeneous in terms of clinical behavior and genetic alterations^5-8. Although the most common mutations in AML are only present in about 30% of adult cases ⁹, patient classification based on mutation profile is a powerful prognostic indicator^6-8. A comprehensive genomic profiling tool was recently developed to identify somatic alterations that affect diagnosis, prognosis and treatment¹⁰. However, genetic heterogeneity means that therapeutic agents to selectively target specific mutations are of limited benefit to most AML patients. Thus, there is also a need to identify non-genetic biochemical commonalities among patients in order to optimize treatment for widespread therapeutic benefit.

1.2. Current Treatments for AML

Sadly, treatment for AML patients has not changed much since the 1970s, despite the wealth of knowledge accumulated since then ^{11, 12}. Standard chemotherapy for patients, the “7+3” combination of cytarabine and daunorubicin infusion, can induce complete remission in approximately 80% of patients under 60 years of age ¹³. Patients above 60 years of age have much lower response rates to standard chemotherapy. If remission is achieved with induction chemotherapy, then consolidation therapy to prevent relapse is administered. Consolidation therapy could consist of additional rounds of high dose chemotherapy or consideration of autologous or allogeneic stem cell transplant. There is debate as to when best to recommend transplant, but high risk patients are often considered for stem cell transplantation in first remission^{14, 15}. Allogeneic hematopoietic stem cell transplantation (HSCT) provides benefit of the graft versus leukemia effect but increases the risk of graft versus host disease and other complications^{11, 14-16}.

Unfortunately these approaches are not curative in most patients, and there has been great interest in developing new AML therapeutic targets and strategies. Particularly encouraging is the introduction of mutational profiling of AML samples and the increase in small molecule inhibitors to selectively target these molecular alterations or other aberrations in AML ^17-19. Many other approaches are underway as well. DNA hypomethylating agents such as azacitidine and decitabine can increase survival in a fraction of older AML patients, but these agents are still being tested in clinical trials and are currently not FDA-approved for the treatment of AML ²⁰. Other agents such as CPX-351 and SGN-CD33A are next-generation formulations of existing chemotherapy and antibody-drug conjugates, respectively, that may prove beneficial for AML patients ²¹.

1.3. Sphingolipid and ceramide metabolism

Sphingolipid metabolism is an up and coming topic in the field of cancer biology, particularly for difficult to treat diseases like AML. Sphingolipids are essential for formation of the plasma membrane and to maintain structural integrity of the cell. However, sphingolipid metabolism also produces bioactive signaling molecules that regulate critical processes like cell survival and differentiation ^22-24. Although sphingolipid metabolism and inter-conversion is complex, ceramide, sphingosine and sphingosine 1-phosphate (S1P) are the bioactive core.

Ceramides are a class of sphingolipids that contain a sphingoid base and a fatty acid. These molecules are synthesized several different ways. De novo synthesis is initiated by condensation of serine and palmitoyl-CoA. Other precursors include glycosphingolipids, sphingomyelin and phosphorylated metabolites like ceramide 1-phosphate and sphingosine 1-phosphate^22-24. The ceramide-S1P flux is most widely studied for its implication in many diseases, including cancer (Figure 1). In the first step of this process, the fatty acid of ceramide is cleaved by a ceramidase to produce sphingosine, which is subsequently phosphorylated by sphingosine kinase to produce S1P. This reversible process is key in regulating cell survival and death²⁵. Dysregulation of these metabolites contributes to the progression of several diseases including multiple cancers ^{22, 26}.

Figure 1. Ceramidase inhibitors reverse the dysregulated sphingolipid rheostat and induce cell death in AML.

The sphingolipid rheostat is imbalanced in AML, with S1P production dominating to produce a pro-survival phenotype. Of the five ceramidases, the mRNA content and enzymatic activity of acid ceramidase (AC) are selectively upregulated in AML. Upon AC inhibition, ceramide accumulates to induce cell death, which reveals a promising therapeutic target for AML.

Ceramide accumulation induces apoptosis and other cell death mechanisms while formation of S1P promotes cell survival ²². In healthy cells, the ratio of ceramide to S1P is relatively stable so that pro-survival and pro-death signals are balanced. However, if this balance is disrupted, cells normally destined for death can proliferate and lead to disease. This balance is tightly regulated by the enzymes involved in formation and breakdown of ceramide. When pro-death signals are prominent, ceramide accumulates via the action of S1P phosphatases that generate sphingosine and ceramide synthases that then produce ceramide. Pro-survival signals dominate in cancer with increased formation of sphingosine and pro-survival S1P through ceramidases and sphingosine kinases, respectively. Exploiting this imbalance in complex diseases like AML provides a unique and promising opportunity to discover essential biochemical dependencies that represent novel therapeutic targets.

1.4. Sphingolipids and AML

In AML, patient cells generally exhibit upregulation of enzymes that degrade ceramide and synthesize S1P^{27, 28}. This suggests a dependence on this pathway for AML blast survival, thus multiple therapeutic approaches to manipulate ceramide levels are at the forefront of current sphingolipid research in AML. For example, treatment of AML cells with sphingosine analog FTY720 rapidly induces ceramide-mediated apoptosis²⁹. The synthetic retinoid fenretinide was shown to induce up to 20-fold increase in ceramide in AML cell lines, yielding cytotoxic effects ³⁰. It has also been shown that treatment with ceramide analog LCL-461 leads to death of AML cells, including those that are drug resistant ³¹. Multiple studies investigate the use of exogenous short chain ceramides as potential therapeutics for cancer, especially in combination with other drugs ³². Combination treatment with C6-ceramide and tamoxifen induces alternations in energy production and decreases expression of anti-apoptotic proteins in AML cells^33-35. Blocking intracellular ceramide modifications as well as treating with exogenous ceramide using nanoliposomes both induced apoptosis in AML ³⁶. S1P-generating enzymes are also important in AML. Inhibition of sphingosine kinase 1 by SKI-I and SKI-178 induces AML cell apoptosis, highlighting the importance of S1P formation in AML cell survival ^{28, 37, 38}.Aberrant signaling induced by the FLT3-ITD mutation represses the production of pro-death ceramide ³¹, thus demonstrating that common molecular alterations in AML may drive sphingolipid dysregulation.

2. Acid ceramidase (AC)

2.1. Ceramidase genes and cancer

There are five ceramidase genes that encode for a family of enzymes whose optimal enzymatic activity is dependent upon pH, which includes acid ceramidase (ASAH1, referred to in this review as AC), neutral ceramidase (ASAH2) and three alkaline ceramidases (ACER1, 2 and 3). AC is synthesized as an inactive precursor that is auto-cleaved to form the α and β subunits of the mature enzyme ³⁹. AC dominates the literature as a therapeutic target, while other ceramidases have not been as extensively studied in this context^{40, 41}. AC is upregulated in multiple cancers including prostate, melanoma and head and neck cancers ⁴², and increased AC leads to proliferation and increased growth rate of oncogenic cells ⁴³. Multiple studies have shown that blocking AC activity induces ceramide accumulation and leads to cell death. These studies suggest that AC inhibition may be a broadly effective strategy for cancers that exhibit sphingolipid dysregulation, particularly those with AC upregulation (Figure 1).

Several publications have shown that elevated AC contributes to cancer progression and demonstrated that targeting AC results in marked improvement in pre-clinical in vitro and in vivo models for solid cancers ^43-46, but few studies have focused on hematological malignancies. Hu et al. demonstrated that indirect targeting, through restoration of the AC repressor IRF-8, decreased AC expression and sensitized chronic myeloid leukemia (CML) cells to Fas ligand-induced apoptosis ⁴⁷. Based on the these studies, AC activity represents a valid target for future AML therapy and supporting evidence is described in the following sections.

2.2. AC as a novel target in AML

Recent work by the Loughran group utilized primary patient samples, cell lines and in vivo murine models to demonstrate AC overexpression and therapeutic potential in AML ²⁷. Based on RNA-Seq data from The Cancer Genome Atlas (TCGA) Research Network, gene expression microarray and AC enzymatic activity screening, AML patient samples exhibit upregulated AC expression and activity relative to normal controls (Figure 2A)²⁷. Interestingly, there was no conclusive evidence that other ceramidases have significant expression or upregulation in the same AML cohorts, with AC expression being at least 10-fold higher than all other ceramidases (Figure 2A). This is a pattern that is currently unique to AML and highlights the importance and specificity of this AC increase. In the TCGA data, average AC expression in 145 AML patient samples was 1.7 times higher than in normal bone marrow cells, while the average AC activity in a separate cohort of 51 AML patient samples was 2.3 times higher than in normal CD34+ controls (Figure 2B). This upregulation explains at least one mechanism for AML cells to enhance their survival by diverting endogenous ceramide towards the production of S1P, which is enabled by the excess sphingosine produced by AC upregulation. S1P, in turn, modulates survival through various mechanisms including receptor dependent or independent signaling pathways ^{48, 49}. Thus, AC is positioned at the fulcrum of the “sphingosine rheostat” between the ceramide and S1P arms where it can play a pivotal role in cell death vs. cell survival determination⁵⁰. Ceramide and S1P can also exert opposite effects on Bcl-2 family members. Multiple groups have shown that ceramide decreases Mcl-1 expression and S1P promotes Mcl-1 expression ^{27, 51-54}. Hence, AC's role in cancer cell survival extends beyond modulating ceramide/S1P level and into regulating the expression of Bcl-2 family members.

Figure 2. AC expression and activity are elevated in AML.

(A) TCGA RNA-Seq gene expression data in 145 AML patients and 5 normal BM samples for ceramidases ASAH1 (acid ceramidase), ASAH2 (neutral ceramidase), ACER1 (alkaline ceramidase 1), ACER2 (alkaline ceramidase 2) and ACER3 (alkaline ceramidase 3). Each bar represents mean and error bars represent standard error of mean (SEM).*, p<0.05, indicates significant difference of AML patient cells compared to normal CD34+ cells (Wald test and Benjamini-Hochberg procedure). (B) AC activity of normal CD34+ cells (n=12, far left, mean ± SEM) and primary AML patient samples (n=51). AC activity is elevated in the vast majority of AML patient cells compared to normal controls (p=0.0016; Wilcoxon rank sum test). Solid line represents the normal mean. Figure reproduced with permission from Tan et al., Oncotarget, 2016.

2.3. Therapeutic potential of AC inhibitors in AML

In the cohort of AML patient samples tested, high AC activity significantly correlated with reduced overall survival and reduced relapse-free survival relative to samples with low AC activity (Figure 3)²⁷. Ongoing studies aim to validate this finding in a larger independent cohort. This correlation suggests that AC activity is clinically relevant, and multiple in vitro and in vivo models were utilized to provide additional evidence that AC is an excellent therapeutic target for AML ²⁷. These studies utilized AC inhibitor LCL204, which has been used previously in prostate cancer and head and neck squamous cancer models ^{55, 56}, and these findings were validated through shRNA-mediated genetic targeting of AC. LCL204 treatment reduced viability and increased apoptosis in AML cell lines and primary patient samples. Pharmacologic and genetic targeting of AC also resulted in decreased Mcl-1 content. Mcl-1 is an interesting target for multiple cancers due to its ability to bind Bak and prevent mitochondrial outer membrane permeabilization ^57-60. Casson et al. showed that combining Bcl-2 family inhibitors with drugs targeting ceramide metabolism effectively killed leukemic cell lines ⁶¹. Hence, targeting AC in AML may have a two-pronged effect through reversal of ceramide/S1P dysregulation and induction of cell death by decreasing Mcl-1 expression. Furthermore, elevated AC expression is selectively found in patient samples, thus the cancerous AML blasts are more sensitive to AC inhibition than normal counterparts²⁷. The therapeutic effect of AC inhibition with LCL-204 was also demonstrated through reduced tumor burden in patient-derived xenotransplanted NSG mice and increased survival in a murine AML model ²⁷. Overall, these studies validate AC as a novel target in AML that could be of great therapeutic benefit, and more AC inhibitors should be tested and studied for clinical use in AML therapy.

Figure 3. High AC activity correlated with lower overall survival and lower relapse-free survival in AML patients.

AC activity values were dichotomized into high/low categories according to whether they were larger than the mean. Patients were filtered to include only those that received standard chemotherapy “7+3”. The main clinical outcomes used were (A) overall survival (OS) and (B) relapse-free survival (RFS). Kaplan-Meier plots and log-rank tests were used to examine the correlation between AC level and OS/RFS. The analysis of RFS was only done for patients who had complete remission (CR) or complete remission with incomplete hematologic recovery (CRi) after primary therapy. Figure reproduced with permission from Tan et al., Oncotarget, 2016.

3. AC inhibitors

3.1. Development and characterization of AC inhibitors

Multiple small molecule AC inhibitors have been developed and are currently being utilized for preclinical studies, however none of these agents have advanced to clinical trials in cancer patients. Methods to improve the synthesis and characterization of AC inhibitors continue to evolve and will facilitate the identification of more efficient compounds ^{62, 63}. Additionally, development of a fluorogenic AC substrate to measure enzyme activity facilitates the identification and characterization of these inhibitors ^{64, 65}. Importantly, a conditional AC knockout mouse has been developed that may be useful to further validate AC as a target in particular cancer types ⁶⁶. This review will focus on new agents since previous reviews detail several of these compounds including ceranibs, derivatives of N-oleoylethanolamide (NOE), and multiple AC-targeting drugs of the LCL series of B13 analogs^67-71.

The LCL series is composed of N-dimethylglycine (DMG)-B13 prodrugs that are metabolized to B13, a known inhibitor of ceramidases. These novel drugs target the lysosome for acid ceramidase specificity. LCL521 was shown to be the most active toward AC and to prevent relapse of prostate cancer in mice ⁷². LCL521 also improved the efficacy of photodynamic therapy in head and neck cancer ⁷³. As described above, ceramide analog LCL204 was shown to decrease AC activity and induce ceramide-mediated apoptosis of AML cell lines and patient samples as well as significantly extend the life of AML engrafted mice ²⁷. This demonstration of pronounced in vivo therapeutic efficacy is particularly important to advance inhibitors towards clinical trials. Another group developed a set of thiol reactive ceramide analogs to inhibit the cysteine hydrolase activity of AC. Two of these compounds, SABRAC and RBM1-12, induced ceramide accumulation followed by cell death in metastatic prostate cancer cells ⁶⁴. Another recent AC inhibitor was suggested as a potent treatment for stage II melanoma. ARN14899 decreased AC activity with an IC₅₀ of 12nM and induced ceramide accumulation. Alone, the compound was only mildly toxic at micromolar concentrations, but viability of proliferative melanoma cells was drastically reduced when combined with chemotherapeutics⁴⁵. Benzoxazolone carboxamide prototypes have also been studied as potent acid ceramidase inhibitors^{74, 75}. Additionally, small peptide targeting of AC with cystatin SA lead to reduced AC activity and increased ceramide accumulation ⁷⁶. This highlights an alternative mechanistic approach to modulate AC activity by targeting the autocatalytic cleavage event that is required to generate the mature active enzyme.

3.2. Clinically used compounds that target AC

Although many studies have shown the potential of therapeutics designed to specifically target AC, identifying FDA-approved agents with AC inhibitory activity is an alternative approach that may have great clinical importance. It was recently discovered that tamoxifen, a commonly used chemotherapeutic and chemopreventive agent, alters sphingolipid metabolism. Both tamoxifen and its metabolite N-desmethyltamoxifen specifically reduce ceramide glycosylation and inhibit acid ceramidase independent of known anti-estrogen mechanisms. Co-treatment with tamoxifen and exogenous ceramide or a ceramide-generating drug induced synergistic apoptosis in AML cells⁷⁷.

Dacarbazine is another approved chemotherapeutic drug with inhibitory activity against AC. Used in the treatment of melanoma, Bedia, C. et al. showed that dacarbazine activates cathepsin B-mediated degradation of AC expression in A375 melanoma cells⁷⁸. Carmofur, a derivative of 5-fluorouracil, was also reported to be an inhibitor of acid ceramidase that worked at nanomolar concentrations⁷⁹. The compound is clinically approved in China and has been shown to strongly inhibit acid ceramidase activity, leading to increased in vivo sensitivity to chemotherapeutics in colorectal cancer⁸⁰. Meta-analysis of three international clinical trials showed modest but significant survival increase in patients treated with carmofur after tumor resection⁸¹. These drugs provide promising opportunities to target AC with more expedited bench-to-patient timelines.

4. Conclusion

The information presented and discussed here shows that dysregulated sphingolipids promote cancer cell survival. AC expression is elevated in AML, which enhances the catabolism of ceramide and promotes the production of S1P. AC inhibition in AML patient samples, cell lines and animal models reduced cell viability and induced leukemic cell death. Hence, AC represents a promising drug target for future development of effective AML therapeutics.

5. Expert opinion

Overall, a growing body of work in cell lines, animal models and primary patient samples strongly supports AC as an essential biochemical activity and promising therapeutic target in AML. Further studies are needed to address open questions and facilitate the leap from bench to bedside. First, there is a need to continue to develop new inhibitory compounds and characterize existing agents. Although many AC inhibitors have been reported, none are approved for clinical use due to various limitations. Such agents will need to have optimal potency, specificity, toxicity and pharmacological characteristics to enable the field to transition from preclinical target validation with tool compounds and genetic knockdown to clinical trials in AML patients.

Second, there is a need to explore differential AC expression and AC inhibitor sensitivity at disease relapse and across the heterogeneous AML subgroups in order to identify the patient population(s) most likely to benefit from these agents. Thus far there is convincing evidence of AC upregulation in the vast majority of AML samples, and further analyses in large independent cohorts are needed to firmly establish the link between AC activity and patient prognosis. Such data could establish AC activity as a prognostic biomarker to identify patients that need more aggressive therapeutic approaches. Additional data and mechanistic studies are required to investigate whether the mutational profile is linked to AC activity and to determine the universality of sensitivity to AC targeting agents. Another aspect of AC upregulation in need of further investigation is within the hierarchy of leukemic stem cells and whether AC confers a survival advantage in these populations. Few studies have addressed primary and secondary resistance mechanisms that may thwart AC targeting agents, which will become important as inhibitors advance to clinical trials in AML and other cancers. All of these efforts would benefit from increased development and application of cell line and murine models engineered to manipulate AC expression and to recapitulate specific genetic alterations that represent AML subgroups. Similar efforts are also required for other sphingolipid metabolic enzymes and ceramide detoxification pathways, such as ceramide synthases, sphingosine kinases and glucosylceramide synthases in order to move the field forward.^{31, 35}

Finally, targeting AC can modulate cellular sphingolipid profiles and impact multiple survival and death pathways. There are multiple reports of sphingolipids' downstream influence on several of the Bcl-2 family of survival proteins, which are also promising targets for therapeutic development^{27, 51-54, 61}. We propose that AC inhibitors may therefore have broader activity than agents that target a single Bcl-2 family protein. Future studies are needed to compare these therapeutic approaches alone and in combination. Additional studies of AC inhibitors are also needed in combination with existing AML therapies and new agents that target molecular alterations in the disease to determine the most promising regimens to advance to clinical trials ^{10, 11, 13}. AC inhibitors may also prove more effective when combined with other sphingolipid modulators such as inhibitors of sphingosine kinases and glucosylceramide synthase, S1P receptor antagonists or exogenous short-chain ceramide nanoliposomes ^{28, 33, 52}. Together these studies will enable the continued advancement of AC inhibits towards clinical usage in a disease that desperately needs more precise and rational therapeutic approaches.

Article Highlights.

• Novel therapeutics for AML are greatly needed due to lack of treatment advances

• Sphingolipid metabolism is dysregulated in AML and contributes to blast survival

• Acid ceramidase is upregulated in AML and elevated activity correlates with reduced patient survival

• Acid ceramidase inhibitors exhibit preclinical efficacy against AML in vitro and in vivo

• Optimized acid ceramidase inhibitors are needed for advancement to clinical trials

Abbreviations

AC

acid ceramidase

AML

acute myeloid leukemia

S1P

sphingosine 1-phosphate